CN110441298A

CN110441298A - Diaper load is determined using color detection or active state

Info

Publication number: CN110441298A
Application number: CN201910372761.2A
Authority: CN
Inventors: B.希弗; T.帕里卡; B.W.吉拉迪
Original assignee: Verily Life Sciences LLC
Current assignee: Verily Life Sciences LLC
Priority date: 2018-05-04
Filing date: 2019-05-06
Publication date: 2019-11-12
Also published as: CN210834672U; EP3787582A1; WO2019213586A1

Abstract

System and method for detecting the volume of the urine in the absorbent commodity for determining such as diaper.Diaper load applies the first measurement for obtaining received environment light from photoelectric detector when light source is closed, and obtains the second measurement from the photoelectric detector when emitting light on absorbent commodity in light source.The application determines the AVHRR NDVI of the light reflected from absorbent commodity by removing ambient light signal from the second measurement based on the first measurement.The application determines the presence of urine in absorbent commodity from AVHRR NDVI.The application also determines the estimation volume of urine in absorbent commodity, and wherein the determination is the active state of the baby based on the elapsed time from the presence of urine and wearing absorbent commodity.

Description

Diaper load is determined using color detection or active state

Technical field

The application generally relates to determine body exudates present in absorbent commodity (such as wearable absorbent commodity) Or the volume of excreta, relate more specifically to variation color and/or inspection based on color change indicator in diaper (diaper) The activity of the diaper wearer measured determines the volume of urine in diaper.

Background technique

The existing solution having for determining body exudates or excreta level in absorbent commodity is not fill Point.For example, some existing solutions are only relied upon using temperature or humidity sensor, this may cause the survey of inaccuracy Amount.For example, humidity sensor is likely located at the position too far from body exudates and can not detect the unexpected increase of humidity.Or Temperature sensor may indicate that raised temperature, but the position sensed may not represent the temperature of absorbent commodity entirety.

Finally, shifting of other solutions of body exudates existing for being used to detect or excreta level due to wearer It moves and is easy to carry out wrong measurement.For example, sensor measurement may include the movement or suction by sensor relative to absorbent commodity Noise or error caused by the movement (for example, the wearer due to absorbent commodity is mobile) of the property received article itself.

Therefore, new solution is at least needed for the above reasons.

Summary of the invention

Present disclose provides a kind of for predicting the load in wearable absorbent commodity (such as body exudates or excretion Object, such as urine) method and system.It has been described in the accompanying claims each aspect of the present invention.

In one example, such as on the computing device the application run determines the urine in the absorbent commodity of such as diaper Volume.When light source is closed, which obtains the first measurement of received environment light from photoelectric detector.When light source is in absorbability When emitting light on article, which obtains the second measurement from the photoelectric detector.Second measurement includes environment light and from absorbability The measurement of the transmitting light of article reflection.Then, using by based on first measurement from second measurement in removal ambient light signal come Determine the AVHRR NDVI of the light reflected from absorbent commodity.The application determines in absorbent commodity according to the AVHRR NDVI The presence of urine.The application also determines the full level of absorbent commodity.The application can also export to be full of determined by instruction The notice of degree.

On the one hand, the application can be based on full level it is determined that replacement absorbent commodity.For example, using that can lead to It crosses and solves the function of the input with instruction elapsed time and active state using regression model to calculate and absorb until expected The time that property article is full of.Using recommended replacement absorbent commodity can be exported before the time in calculating or the time in calculating Notice or warning.

On the one hand, determine that full level is the size based on absorbent commodity.

On the one hand, determine full level include use one in (i) regression model or (ii) machine learning model or Multiple functions to solve the input with instruction elapsed time and active state.

On the one hand, absorbent commodity is the diaper of baby's wearing, and determines that full level is also based on (iii) since suction The property received article is placed on the time additionally gone through risen with baby.

On the one hand, active state include (i) it is sleeping or (ii) it is awake in one.

On the one hand, absorbent commodity is the diaper of baby's wearing, and the application further includes one accessed about baby Group demographics (demographics), and determine that full level is based in part on the demographics.

On the one hand, the type of application access absorbent commodity.The type is (i) night diaper or (ii) in the daytime in diaper One kind, and determine full level be based in part on the type.

On the one hand, determine that active state includes that three-dimensional inertia measurement is received from movable sensor for a period of time.The determination Including calculating the statistical data derived from inertia measurement.The determination further includes that inertia measurement and statistical data are supplied to prediction mould Type.The determination further includes activity according to prediction model and determining based on inertia measurement reception.

On the one hand, the second measurement includes the independent measurement of feux rouges, green light and blue light.The AVHRR NDVI for determining light includes Determine the independent intensity of feux rouges, green light and blue light.Determine that full level depends on the intensity of feux rouges, green light and blue light.

On the one hand, light source is configured as the light of transmitting specific wavelength, which is based on absorbent commodity to not The responsiveness of the light of co-wavelength and determination.

On the one hand, absorbent commodity includes containing the printing or coating zone for including color change indicator.

On the one hand, the application also obtains the color calibration value of storage from memory.This applies the normalization based on light The color calibration value of measurement and storage determines the color of absorbent commodity.Stored color calibration is determined using white object Value, and determine that color includes white level correction.

On the other hand, a kind of system for determining the volume of the body exudates in absorbent commodity include light source, Photoelectric detector, movable sensor and processor.Processor is configured as being received when light source is closed from photoelectric detector Environment light first measurement.Processor is configured as obtaining the second measurement from the photoelectric detector when source emissioning light.The Two measurements include the measurement of environment light and the transmitting light reflected from absorbent commodity.Processor is configured as by based on environment light First measurement from second measurement in removal ambient light signal come determine from absorbent commodity reflect light AVHRR NDVI.Place Reason device is additionally configured to be obtained from movable sensor including the inertia measurement for the period each of one group of period Time series data.Processor is additionally configured to send the AVHRR NDVI or (ii) time series number of (i) light to external equipment At least one of according to.Processor is additionally configured to receive the full level of absorbent commodity from external equipment.

On the one hand, described to send so that external equipment is measured existing for urine by using regression model with being based on self-test Function is solved by the active state of time and the baby of wearing absorbent commodity to determine until expected absorbent commodity is full of Time.

On the one hand, described to send so that external equipment is risen based on (iii) since absorbent commodity is placed on baby Time for additionally going through determine full level.

On the other hand, a kind of system include store non-transitory computer executable instructions computer-readable medium and The computer-readable medium is communicably coupled to for executing the processing equipment of the non-transitory computer executable instructions.It executes Non-transitory computer executable instructions configure the processing equipment to execute operation.The operation is included in when light source is closed from photoelectricity Detector obtains the first measurement of received environment light.Is obtained from the photoelectric detector when operation is included in source emissioning light Two measurements.Second measurement includes the measurement of environment light and the transmitting light reflected from absorbent commodity.The operation includes by being based on First measurement of environment light removes ambient light signal from the second measurement to determine the normalization of the light reflected from absorbent commodity Measurement.The operation includes that the presence of urine in absorbent commodity is determined according to the AVHRR NDVI of light.The operation includes being based on (i) One or more of the active state of baby of elapsed time and (ii) wearing absorbent commodity is come since urine exists Calculating absorbent commodity should be replaced.The calculating is based on following one or more: being placed on baby from absorbent commodity Play the full level of elapsed time and (ii) absorbent commodity.

On the one hand, described calculate includes using one or more of (i) regression model or (ii) machine learning model Function is solved based on elapsed time and active state.

On the one hand, active state is determined by operating as follows: receiving one group of three-dimensional inertia measurement from movable sensor For a period of time, the statistical data derived from inertia measurement is calculated, to prediction model offer inertia measurement and statistical data, and according to Prediction model and the activity that determination is received based on inertia measurement.

It refers to that these illustrated examples are not limited to or limit the scope of the present disclosure, and is to provide example to help to manage Solve the disclosure.Illustrated examples are discussed in a specific embodiment, provide further description.By consulting this explanation Book, it will be further appreciated that the advantages of various examples provide.

Detailed description of the invention

When reference attached drawing reads following specific embodiments, it is better understood with the feature of the disclosure, embodiment and excellent Point.

Fig. 1 depicts the exemplary block diagram of baby's sensing system according to some aspects of the disclosure.

Fig. 2 depicts the exemplary block diagram of the Color Detecting System of some aspects according to the disclosure.

Fig. 3 depicts the absorbability object with color change indicator and sensor device of some aspects according to the disclosure Product.

Fig. 4 includes Fig. 4 A and 4B according to some aspects of the disclosure.Fig. 4 A indicates the sensor cloth of sensor encapsulation 400 The exemplary top side view of office.Fig. 4 B indicates the exemplary underneath side elevation view of the sensor placement of sensor encapsulation 400.

Fig. 5 is depicted to be configured according to the exemplary color detector cell of some aspects of the disclosure.

Fig. 6 is the flow chart for describing the method for the detection color according to some aspects of the disclosure.

Fig. 7 is the stream for being used to determine movable illustrative methods according to movable sensor according to some aspects of the disclosure Cheng Tu.

Fig. 8 is according to some aspects of the disclosure for being lived by using prediction model to be determined according to movable sensor The flow chart of dynamic illustrative methods.

Fig. 9 is the side of the volume of the body exudates in the detection absorbent commodity described according to some aspects of the disclosure The flow chart of method.

Figure 10 is described according to some aspects of the disclosure for executing and the detection of color detection and body exudates The figure of the exemplary computing system of relevant function.

Specific embodiment

Aspects described herein are provided by using Color Detecting System combining movement categorizing system and are accurately determined The solution of the volume of body exudates or excreta (for example, urine or excrement) in absorbent commodity.Absorbent commodity Preferably wearable structure, such as baby diapers etc..More specifically, light-pulse generator can be used in exemplary color detection system Accurately detect the color of the object of color change indicator in such as absorbent commodity (for example, diaper) etc.Color inspection Examining system can compensate or adjust its operation to minimize the influence of environment light, so that this can be used there are environment light System.Activity classification system can be worn by analyzing from the measurement that the inertial sensor for being attached to absorbent commodity obtains to determine Wear the active state of the baby of absorbent commodity.

More specifically, exemplary color detection system includes one or more light sources (such as LED), is configured to detection light The electronic circuit or equipment of one or more photoelectric detectors and such as photometer front end or general processor, it is described general Processor can be configured to receive the information about the color detected, the contribution of filtering environmental light, and export what instruction detected The signal (for example, as diaper condition information, for determining or controlling further operation) of color.Color Detecting System is attached It is connected in baby diapers and Color Detecting System is adapted to allow for shine in a part of the diaper comprising color change indicator. Color change indicator can with for example, based on substance to be detected existence or non-existence (such as it is exemplified here in body outflow Object or excreta) change color.

Example activities categorizing system is received from the movable sensor of such as accelerometer or gyroscope and is measured.Sensor is placed With wearer, such as fixed on absorbent commodity (such as by being clipped on diaper).By using prediction model or shape State machine, activity classification system determine that wearer is awake or sleeping.Wearer is awake or sleeping combine such as is begged for above Other data of the diaper condition information of opinion etc improve the presence of the body exudates of detection such as urine or the standard of volume True property and reliability.Therefore, system disclosed herein is provided better than the color only relied upon in detection diaper or another sensor The advantages of system of the color of change indicators, to help to correct or avoid the shifting of the wearer by sensor or sensor Mistake caused by dynamic.

These illustrated examples are provided to introduce general theme discussed in this article to reader, and the present disclosure is not limited to this Example.The example of color detection following section describes various other non-limiting examples and using pulsed light.

Fig. 1 depicts the exemplary block diagram of baby's sensing system according to some aspects of the disclosure.Fig. 1 depicts baby Sensing system 100 comprising color sensor 120, movable sensor 130 and microcontroller 101.Microcontroller 101 includes urine Piece load applies 116 and data 117 using 111, activity classification using 102, color detection.Data 117 may include demographic data (gender, age, weight etc.), about data, active state of baby of quality of materials of absorbent commodity etc..Data 117 can To be inputted via user interface, for example, inputting from nursing staff or being downloaded from external equipment.

In one example, diaper load is grasped using 111 and activity classification using 116 using 102 with color detection together Make the volume to determine the body exudates of such as urine in absorbent commodity 160.As further explained herein, absorbability object Product 160 may include the color-bar 161 that presence to urine or amount have reaction.In one example, color detection is using 111 operations To determine the color change in color bar 161.In another example, color detection can be used for detection such as using 111 Presence based on another chemical substance change any other object of color change indicator of color etc color (for example, The output instruction color detected rather than the signal of color change detected).Color detection application can be configured further Urine in absorbent commodity 160 (or other chemical substances) are determined at the color detected or the color change detected is used Presence.Color detection application can export signal existing for instruction urine.Signal can be binary, or can be base In the classification score of the color variation amplitude of the color detected.

The state for the baby that activity classification can determine wearing absorbent commodity 160 using 116.It can be by movable sensor 130 (can be accelerometer, gyroscope or other sensors type) are placed or are adhered on absorbent commodity 160.Activity point The state that class can determine baby using 116, such as baby is to fall asleep, regain consciousness or rest etc..

In one example, movable sensor 130 is attached to the clothes or absorbent commodity 160 of baby.It may include accelerating The movable sensor 130 of degree meter or gyroscope provides measurement.Activity classification can be received from movable sensor 130 using 116 and be surveyed It measures and uses such as machine learning model, the prediction model of state flow model or algorithm to determine by the baby of wearing movable sensor The activity that youngster carries out.In one example, the training prediction model is with the mobile determining ongoing work of baby based on baby It is dynamic, it such as sleeps or stays up late.The prediction model can be trained by the data collected for individual or group of individuals.

For example, prediction model can receive the training data for representing particular individual.Training data may include being passed by mobile The individual that sensor 130 determines is mobile and relevant brass tacks (ground truth) is (by the specified correct work of training label It is dynamic).Therefore, when receiving new data from movable sensor 130, trained prediction model foresight activity.In another example, Prediction model can be trained based on the training data collected from one group of individual, which does not need to include predicting it later Movable individual.Therefore, prediction model can learn from the record behavior of particular individual or one group of individual, and use study The activity of the individual or another individual is predicted in behavior.

Then, activity classification system can export the foresight activity information (such as baby is in deep sleep) of instruction baby Signal, for determining or the further operation of control.The suitable procedure for activity classification is described with reference to Fig. 7 and 8 Example.

Using may include from color detection using 111 urine existing instruction diaper condition information and can wrap Include from activity classification using 116 foresight activity state foresight activity information, diaper load can determine absorption using 102 Volume of urine present in property article 160.Volume of urine is properly termed as diaper load.In some cases, statistics can be used Method.Different factors can be used, the shape of elapsed time, diaper size or baby such as since first time urination event State.The example for determining the process of the volume of urine in absorbent commodity is described referring to Fig. 9.

Baby's sensing system 100 may be embodied as example, for being attached on the slim material of such as plastics or flexible substrate To the attachment of wearable items.For example, baby's sensor-based system 100 can be 1-2 centimetres wide and 2-5 millimeters thick.As referred to Fig. 3 institute It discusses, baby's sensing system 100 is made to sufficiently small and thin to be placed in the absorbent commodity of such as diaper.One In a example, baby's sensing system 100 can be placed in the diaper including color change indicator, so that light source and photoelectricity inspection Device is surveyed to be aligned with color change indicator.

In one example, baby's sensing system 100 can be distributed in attachment and external equipment fixed to absorbent commodity Between.For example, the operation that microcontroller 110 executes can be executed by external equipment.Therefore attachment may include communication module, example Such as transceiver, is used for and external device communication.Example includes by local wireless electrical connection (for example, WiFi or bluetooth) connection Monitor and be connected to the remote server of microcontroller 101 via internet.For example, diaper load applies 102, color sense External equipment can be entrusted to for some or all operations described herein using 116 using 111 and/or activity classification by surveying.It is excellent Point includes but is not limited to the battery life for reducing baby's sensing system 100, or due to additional data set (for example, training data) Or the processing capacity on external equipment availability and improved performance.

Any or all process described herein, such as the process discussed about Fig. 6-9, can be held by external equipment Row.For example, microcontroller 101 is by the measurement color such as from color sensor 120 or the movement from movable sensor 130 The data of data are sent to external equipment.Then, external device processes operation is with the significant of the color that confirmly detects or movement Property (significance).

Fig. 2 depicts the exemplary block diagram of the color detection system according to some aspects of the disclosure.Fig. 2 includes color sense Examining system 200 comprising light source 202, photoelectric detector 204, processor 206 and microcontroller 101.Microcontroller 101 can be with Realize the function of color detection system 200, baby's sensing system 100 or both.In addition, for illustrative purposes, in color Microcontroller 101 is depicted in sensing system 200, but different microcontroller, microprocessor or other processing can be used Device.On the one hand, there is only one in processor 206 and microcontroller 101.

Color detection system 200 can be configured as measurement object 251 (for example, in absorbent commodity or absorbent commodity Color-bar) color to determine the load of absorbent commodity.The system can obtain with environment light there are unrelated colors Measurement for example, environment and compensated to it by measurement, or is reduced or minimized by adjusting (such as tuning) light source Influence of the environment light to the signal detected.

Color detection system 200 further includes microcontroller 101.Microcontroller 101 can be any controller, processor, Specific integrated circuit or other processing equipments.The example for calculating equipment is shown in Figure 10.Microcontroller 101 can execute color Sensing is using 111 and other processor-executable instructions to execute all aspects of this disclosure.

The function of microcontroller 101 can realize that vice versa by processor 206.Microcontroller 101 can store data 117, data 117 may include the state of baby, the demographic information about baby, the specific absorption dressed about baby The information etc. of property article.

Environment light 250 can be any kind of light not generated by light source 202 existed in the environment, may include Such as the light from natural source of sunlight, or such as via incandescent source, halogen light source, light emitting diode (" LED ") The artificial light of the light of the generations such as light source, fluorescent light source, laser source.Even if environment light (multiple) environment light source according to existing for can be with With different chromatographies, baby's sensor-based system 100 can also electronically remove what this environment light detected photoelectric detector The contribution of light, and accurately detect based on the reflected light from light source 202 color of object 251.

Light source 202 includes that can operate so that light to be radiated to one or more light sources on object 251.Light source can be basis Any suitable artificial light sources of the disclosure, including LED, incandescent source or other light sources.Multiple discrete light sources can be individually real It is existing, or realized via by multiple independent combination of light sources at the integration packaging of single source.

Light from light source 202 can be generated in one or more certain wave strong points, or may include multiple wavelength.In In one example, there are three light sources for the tool of light source 202: the green light and wave that wavelength is the feux rouges of 623 nanometers (" nm "), wavelength is 523nm The blue light of a length of 455nm.Color change depending on application, the expection color gamut of target object or such as litmus paper item refers to Show device, expected ambient light spectrum or any other suitable factor, the light of other wavelength can be used according to different examples.In In some examples, light source can be it is tunable, to allow to select one or more wavelength that there is small contribution to environment light Light.For example, if the locally or globally minimum radius of the environment light instruction first wave strong point detected by photoelectric detector, baby Youngster's sensing system 100 can be with tuning source 202 substantially to be shone with first wave length.

In this example, Color Detecting System is by activation short duration (for example, 1-5 microsecond is to 500 milliseconds, referred to as " pulse width ") deactivate light source then to make the light pulse of the sending of light source 202.Any suitable pulse width can be used for specific Using.Light source 202 can generate individual pulse for red, blue and green, and export corresponding value.For example, the arteries and veins of 5 microseconds Rush the color that width is advantageously possible for detection color change indicator.Short pulse width enables baby's sensing system 100 with that This quick continuous mode chopping and the light for detecting different colours (such as red, green and blue).

The ambiguity of the type for using the light for enabling color detection system 200 to eliminate object reflection of pulsed light.Specifically Ground, color detection system 200 can be from including detecting from the detection light of the light of 202 chopping of light source and filtering environment light.In In some examples, color detection system 200 can at regular intervals (for example, every ten minutes) or in response to event (such as with Button on Color Detecting System is pressed at family or humidity sensor detects that humidity level is more than threshold value) to 202 chopping of light source. In addition, the use of pulsed light can reduce the power consumption of color detection system 200, to increase color detection compared with constant light The time quantum that system 200 can be operated according to battery.

When light source 202 is by chopping, the detection light at photoelectric detector 204 can be environment light 250 and come since right As the combination of the light of the light-pulse generator 202 of 251 reflections.When light source 202 is inactive, the light that photoelectric detector 204 detects is Environment light.By carrying out chopping to light source 202, color detection system 200 can obtain the baseline about ambient light spectrum first Information, so that Color Detecting System can filter out received light when 202 activity of light source.Chopping also allows baby to sense system System 100 is by deactivating light source 202 when not carrying out color measuring come energy saving.

Photoelectric detector 204 receives light, and (including the light reflected from object 251, either environment light is still sent out by light source 202 The light penetrated), and based on a received photogenerated sensor signal.Photoelectric detector 204, which can be, is able to detect and measures light Any equipment, such as photodiode, phototransistor, complementary metal oxide semiconductor (CMOS) imaging sensor, electricity Lotus coupled apparatus (CCD) sensor or photo-resistor.

Photoelectric detector 204 can detecte wide spectrum, and export the information for the light that instruction detects.For example, Photoelectric Detection Device 204 can produce the electricity output proportional to the wavelength of light is received.Photoelectric detector 204 can provide the three of RGB triple A output, for example, corresponding to red value, another value for green and another value for blue.

More specifically, the value of triple corresponds to the amplitude of the light in wave-length coverage corresponding with particular color.Therefore, First value to receive light in red amplitude it is proportional, second value with reception light in green amplitude it is proportional, third value with connect The blue amplitude received in light is proportional.

On the one hand, photoelectric detector 204 can be the array of individual photoelectric detector.Each photoelectric detector can match It is set to the color of measurement light.Such as the red second photoelectric detector measurement blue of a photoelectric detector measurement, third light Photodetector measurement green.

Processor 206 is the electronic circuit or equipment of such as general processor.Processor 206 can be in analog domain, number It is operated in domain or both.Processor 206 can distinguish the true colors of object 251 independently of any environment light.Processing Device 206 receives the first output for indicating environment light, such as the output collected when the closing of light source 202 from photoelectric detector 204.Work as light When source 202 is by chopping, processor 206 receives the second output from photoelectric detector 204.Processor 206 identification first output and Difference between second output is especially the color of the reflected light on object from pulsed light separation to separate the color of object.

On the one hand, processor 206 receives the level of the intensity of the broad-spectrum light of instruction expression environment light (that is, light source 202 closes The time point closed) and the second time point of instruction intensity level, at the second time point, red, blue and green three kinds of face Color first is that chopping.Then, processor 206 can be by comparing the intensity of environment light and the intensity of pulse color Relatively carry out to eliminate the contribution of pulse color from environment light.

First group of red, green that processor 206 receives the time point closed for light source 202 from photoelectric detector 204 With blue level and light source 202 by second group of red, green and blue level at the second time point of chopping.Processor 206 calculate the difference between first time points and red level between the second time point, to calculate from the red of pulsed light The contribution of color, green and blue level.

Processor 206 can be special photometer front end, such as Analog ADPD105、ADPD106 Or ADPD107.Processor 206 can be configured as activating light source 202 and measure by 204 received signal of photoelectric detector.Example Such as, processor 206 can receive simulation input from photoelectric detector 204, by using analog-digital converter (ADC) by simulation input Be converted to numeral output, then will indicate that the numerical value of color detected is stored in internal storage, so as to later with it is another A value is compared.

By this method, processor 206 can be configured to eliminate the contribution of analog domain ambient light 250, and export instruction pair As the analog signal or digital value of 251 color.For example, processor 206 can provide the RGB for such as indicating the color of object 251 The output of triplet.

On the one hand, processor 206 can have multiple sense channels, and each sense channel corresponds to light source 202 and light Photodetector 204 is right.As further described about Fig. 4 A and 4B, each channel can be exclusively used in specific light source-photoelectricity inspection Survey device pair or " unit ".Each unit can be physically separate, and processor 206 is allowed to measure color in multiple positions. Processor 206 can also be with differently light of the chopping from discrete cell of the light from another unit.

Color detection using 111 can for from 202 received signal of light source provide such as calibrate or white balance additional function Energy.For example, microcontroller 101 receives the numeral input of the color of reception light of the instruction from processor 206.Numeral input can be with Including red, green and blue level.It can be tone by red, green and blue level conversion that color detection, which applies 111, full With degree and brightness/value, and calculating is executed to tone, saturation degree and brightness/value.

Color detection can also calibrate received color value using 111.For example, color detection using 111 can for example when As known to the known color that white or gray card indicate is retrieval such as detected value when being presented to the object of photoelectric detector 204 Value.Color detection can adjust received red, blue and green level using 111 according to known calibration value.

On the one hand, microcontroller 101 may be coupled to transceiver 212.Transceiver 212 can according to such as bluetooth, Any suitable wireless protocols of WiFi, near-field communication etc. are communicated.Using transceiver 212, microcontroller 101 can be transmitted The color of object 251, or if detecting the body exudates in absorbent commodity, notify external equipment absorbent commodity It has been dirty.Microcontroller 101 can transmit information to the remote of such as smart phone, smartwatch or other wearable devices Journey equipment, or be transmitted to the remote computer of such as server (such as server based on cloud), to be further processed and Analysis.

Microcontroller 101 can be transmitted to remote service from processor 206 via the color that transceiver 212 will test The value for indicating the expection reflection colour from object can be mapped to body present in absorbent commodity by device, remote server The prediction volume of effluent.This mapping can be completed by table.For example, table may include red-blue-green (RGB) ternary Mapping between the prediction volume of group or triple range and body exudates.

Object 251 can be color change indicator or presence based on chemical substance and change the other materials of color. On the one hand, color change indicator can dissolve in the presence of the liquid of such as urine.Therefore, baby's sensing system 100 can To detect the disappearance of the variation of color, the appearance or color of color.

Such as and as further discussed about Fig. 3-6, in one application, baby's sensing system 100 is for passing through Read the presence that body exudates are measured based on the color change indicator of the presence of liquid or stereomutation color.It is exemplary Color change indicator include pH item or change the litmus paper item of color based on the pH level that detects.Color detection system 200 simultaneously, will determine the amount of the pulsed light reflected in light pulse to color change indicator.

More specifically, microcontroller 101 is utilized to program, institute from the data point of one or more wavelength absorption curves The acidity or pH that wavelength absorption curve is stated corresponding to different level are horizontal.By by the extinction level of the light of specific wavelength with it is specific Acidity level match, microcontroller 101 can determine the particular liquid of such as body exudates volume or specific pH It is horizontal.For example, if the absorbance of measurement is 0.1, microcontroller 101 determines existing liquid for the light of 440nm wavelength It is preliminary, and exist with low volume.In another example, if the absorbance of the 440nm light of measurement is 0.3, due to High-caliber acidity, microcontroller determine liquid and exist with high volume.In this way, microcontroller 101 does not need to calculate Intermediate pH is horizontal, but absorbance or reflectivity can be mapped directly to the volume of body exudates.Microcontroller 101 can be with Expected reflectivity is determined based on the absorbance of specific frequency, that is, is expected the specific frequency measured by photoelectric detector 204 Light quantity.

Microcontroller 101 can retrieve the calibration value of storage from memory, and determined and absorbed according to color and calibration value The amount of body exudates present in property article.For example, microcontroller 101 can store and be mapped to given value or color gamut The corresponding amount of existing body exudates or the table of volume.Microcontroller 101 can have multiple tables, for example, a table Lattice are for each of a different set of color change indicator.In addition, for example, to use different color changes to indicate In the case where device, the table can be updated.

The wavelength of light source 202 can be changed based on specific application or color change indicator.For example, pH color change refers to Show that device there can be bigger response in certain wave strong point, and therefore can choose or tuning source 202 is to emit this wavelength Light.In this way, by using have can be by the light source for the specific wavelength that color change indicator preferably reflects, system Stronger reflected impulse optical signal can be received from object.This can permit system more accurately determine object color, and because This more accurately determines pH value or corresponding volume based on identified color.When the color value of color change indicator is not with pH Variation linear change when, it is this accurately may be especially valuable.

Fig. 3 depicts the suction with pH sensitive colors change indicators and sensor device of some aspects according to the disclosure The property received article.Fig. 3 depicts absorptive article system 300 comprising absorbent commodity 301, sensor encapsulation 310 and color-bar 351.In this example, baby's sensing system 100 of Fig. 1 is implemented in sensor encapsulation 310.In addition, in some instances, it is more Color detection system or multiple light sources for single polychrome detection system and photoelectric detector can be deployed in absorbent commodity Different location, preferably to detect the presence of the body exudates in absorbent commodity at multiple and different positions.

Color-bar 351 is shown as extending downwardly and (being shown with striped) to the other end from one end along the centre of absorbent commodity. Because body exudates can be unevenly distributed in absorbent commodity, color-bar 351 is placed on downwards absorbability The middle position of article increases the chance for the body exudates that color-bar 351 will test in absorbent commodity 301.But it is colored Item 351 can be located at the different zones of absorbent commodity.For example, color-bar 351 can be located at the front of absorbent commodity, or Positioned at the edge of absorbent commodity 301 or any combination of these or other positions.

As can be seen that sensor encapsulation 310 is aligned with color-bar 351, so that light source and photoelectric detector components are located in On color-bar 351.In some instances, sensor encapsulation 310 can be removed from absorbent commodity 301.For example, sensor encapsulates 310 can adhere on absorbent commodity 301, to prevent 310 sliding of sensor encapsulation, while allow its removal.

Absorbent commodity 301 can be any suitable absorbent commodity, and such as common disposable urine napkin repeatable makes Cloth diaper, Pantiliner, adult diaper etc..Color-bar 351 is color change indicator, be designed to respond in It contacts the substance with special properties (such as pH is horizontal) and changes color.For example, color-bar 351 can be bromocresol green, The pH for the liquid being exposed to based on color change indicator and change color.The color of bromocresol green item is with detecting The pH value of body exudates and change.Other color change indicators can be used.The pH level detected can be with body stream The volume of object is related out, because pH level changes with the volume change of body exudates in absorbent commodity.Therefore, it searches Table or function can be used for determining the volume of given pH level or the color of color change indicator.

Sensor encapsulation 310 may include baby's sensing system 100 and/or color detection system 200, may include soft In the impermeable encapsulation of property.For example, sensor encapsulation 310 has the shell for being able to bear body exudates and excrement, and And it is sufficiently thin so that the discomfort of the wearer of absorbent commodity will not be caused.Sensor encapsulation 310 can be used such as thin The flexible substrate of plastics, fluorubber or silicon substrate thermoplastic sulfurized rubber (tpsiv) manufactures.

Sensor encapsulation 310 can be placed in absorbent commodity in a variety of different ways.On the one hand, sensor encapsulates 310 can be removed and be inserted into new absorbent commodity.Sensor encapsulation 310 can with capable of washing or erasable material or Pouch covering.For example, sensor encapsulation 310 can be inserted in absorbent commodity or adhere to the inside of absorbent commodity.Sensing Device encapsulation 310 also can be inserted in the pocket in absorbent commodity or pouch.This pocket or pouch can be it is hermetically sealed, For example, being encapsulated in the transparent plastic for allowing light to pass through.Sensor encapsulation 310 can also be permanently attached to absorbent commodity In and abandon after a single use.Sensor encapsulation 310 can also adhere to absorbability via velcro fastener or similar material The outside of article.

Fig. 4 A and 4B depict according to some aspects of the disclosure can be placed in absorbent commodity or its outer surface On sensing system exemplary layout.Fig. 4 A indicates the exemplary top view of the sensor placement of sensor encapsulation 400.Figure 4B indicates the exemplary view from bottom to top of the sensor placement of sensor encapsulation 400.Sensor encapsulation 400 can be with Fig. 3 The absorbent commodity 301 of description is used in combination for example to execute with color detection discussed above using associated function.It moves Dynamic sensor (not shown) may include in sensor encapsulation 400 or being provided separately, to provide and activity point discussed above Class applies associated function.

As shown, bottom is the side for being positioned facing color-bar 351 and being aligned with color-bar 351.Biography shown in Fig. 4 Sensor system, can be with by the color of color change indicator in detection absorbent commodity when being placed in absorbent commodity In conjunction with the built-in system of such as microcontroller 101 of color mapping to body exudates volume being determined in absorbent commodity The presence and volume of existing body exudates.

Sensor encapsulation 400 includes battery 402 and one or more color detector unit 420a-n.Sensor encapsulation 400 can also include 404, two electric connector 440-441 of switch, easily diverging organic compound (" VOC ") sensor 410, temperature Spend sensor 411, humidity sensor 412, additional ambient light sensor 414, processor 206, microcontroller 101 or transceiver 212.Additional ambient light sensor 414 can be used in combination with photoelectric detector, to improve or enhance the light of sensor encapsulation 400 Detectability.Some aspects can not include above-mentioned all components, or including its modification.

In addition, sensor encapsulation 400 can be caused based on the threshold level of the body exudates detected it is such as audible Beeping sound alarm.Therefore, sensor encapsulation 400 may include loudspeaker or other audio output apparatus.Sensor encapsulation 400 Alarm can also be made to be transmitted to another equipment for example operated by watchman.On the other hand, sensor encapsulation 400 can be to Another equipment transmits alarm.Sensor encapsulation 400 may include the transmitter that radio signal can be transmitted to external equipment Or transceiver.The color detection operated on microcontroller 101 can also be by event (for example, when detecting body using 111 Effluent) memory is recorded to send nursing staff to later.

Sensor encapsulation 400 may include one or more color detector unit 420a-n.For example, multiple color detectors Unit 420a-n can increase the ability for the variation that sensor encapsulation 400 is detected across the body exudates of absorbent commodity.Cause It may be unevenly distributed in absorbent commodity for body exudates, so the color of color-bar 351 may not be along face The length of color change indicator equably changes.In addition, the presence of multiple color detector unit 420a-n makes it possible to calculate Multiple data points are more accurately to estimate total load.

Each color detector unit 420a-n includes the light source of such as LED and the Photoelectric Detection of such as photodiode Device.In some respects, as further discussed about Fig. 4, color detector unit may include multiple light sources or the inspection of multiple photoelectricity Survey device.Each color detector unit 420a-n detects the light reflected by the object 251 of such as color bar 351, such as environment light Or come from the pulsed light of (multiple) light source.The output of each color detector unit 420a-n is provided to processor 206.Processing The output of device 206 can be supplied to microcontroller 101.In some instances, each color detector unit 420a-n can have There is application specific processor 206, and in some instances, multiple color detector unit 420a-n may be coupled to common processor.

Sensor encapsulation 400 may include switch 404 to activate or deactivate sensor encapsulation 400.Switch 404, which can be, appoints What is suitably switched, such as battery 402 is connected to the sensing of such as color detector unit 420a-n and sensor 410-414 The rocker-type on/off switch of electronic device in device encapsulation 400.Switch 404 is also possible to button switch, whithin a period of time Activate the electric power of the encapsulation 400 from battery 402 to sensor.Sensor encapsulation 400 is configurable to be automatically closed to save battery Electricity.On the one hand, in conjunction with microcontroller 101, sensor encapsulation can be remotely activated.For example, user can be ordered by voice Prompt external equipment is enabled, this makes external equipment via the state being wirelessly connected to the transmission of microcontroller 101 to absorbent commodity Request or open or close sensor encapsulation 400 request.

Sensor encapsulation 400 may include one or more electric connectors of such as electric connector 440-441.Electric connector 440 and 441, which can be used for debugging sensor encapsulation 400, calibration sensor encapsulates 400, sensor encapsulation 400 is reset to factory sets It sets, upgrade software etc. in sensor encapsulation 400.

As discussed with respect to FIG. 1, processor 206 can distinguish the color of such as object of color change indicator.It is micro- Controller 101 can execute the application that such as color detection applies 111, can execute the calibration of the color value detected.Such as Fruit sensor encapsulation 400 detects that there are body exudates in absorbent commodity, then transceiver 212 can notify external equipment.

On the one hand, sensor encapsulation 400 can also include VOC sensor 410.VOC sensor 410 can be according to bowl-shape The presence of the easy diverging organic compound of the detection of VOC present in mobile or blood such as excrement.In conjunction with from color detector list The data that first 420a-n is obtained, VOC sensor 410 can be based on one or more easy diverging organic compounds detected to micro- Controller 101 provides additional information.

On the one hand, sensor encapsulation 400 can also include temperature sensor 411.Temperature sensor 411 can be from such as The substance of body exudates detects heat.In conjunction with the data obtained from color detector unit 420a-n, temperature sensor 411 can To provide the additional information that such as temperature temporarily increases etc to microcontroller 101.Because temporarily raised notice can be with for temperature There are body exudates for instruction, therefore the accuracy and reliability of detection can be improved in this information.

On the other hand, sensor encapsulation 400 may also include humidity sensor 412.Humidity sensor 412 can such as root According to the presence of body exudates detection humidity.In conjunction with the data obtained from color detector unit 420a-n, humidity sensor 412 Additional information, such as temporary raised notice of humidity can be provided to microcontroller 101.Because humidity is temporarily increased and can be indicated There are body exudates, therefore the accuracy and reliability of detection can be improved in this information.

On the other hand, sensor encapsulation 400 can also include additional ambient light sensor 414.As shown, can put Additional ambient light sensor 414 is set, color detector unit 420a-n is made it away from, more accurately to detect environment light.In conjunction with The data obtained from color detector unit 420a-n, adding ambient light sensor 414 can be additional to the offer of microcontroller 101 Information allows microcontroller 101 preferably to eliminate contribution of the environment light to the color of color change indicator.Additional environment light Sensor 414 can also be provided to microcontroller 101 about the absorbent commodity for dressing wherein placement sensor encapsulation 400 Baby whether the information in darkroom.It is beaten for example, sensor encapsulation 400 can provide the lamp in baby room to nursing staff Open or close the instruction or notice closed.

As discussed, sensor encapsulation 400 may include multiple color detector unit 420a-n.There are more than one Color detector unit 420a-n allows to improve accuracy and reliability.For example, a color detector unit 420a-n may It is blocked by object, so that the detected value from the unit is unavailable, or because body exudates may be unevenly distributed over It in absorbent commodity, and is therefore unevenly distributed on color change indicator, uses more than one color detector list A possibility that first 420a-n increase in color detector unit 420a-n one detects body exudates.With this side Formula, breaks down in any one of color detector unit 420a-n or in the case where misalignment, additional color detection Device unit 420a-n helps to increase robustness.In addition, additional more multiple-unit 420a-n, which can be provided, can help to estimate always The additional local information of load.On the contrary, less color detector unit 420a-n can simplify whole system framework and go back Power consumption can be reduced.

In another example, in the system in tool there are three detector cell 420a-c, if a detector cell 420a is returned and detector cell 420b and 420c inconsistent color measuring, then microcontroller 101, which can be ignored, carrys out self-test The measurement of device unit 420a.

Fig. 5 is depicted to be configured according to the exemplary color detector cell of some aspects of the disclosure.As discussed, all If the sensing system of sensor encapsulation 400 includes one or more color detector unit 420a-n.Fig. 5 is illustrated in greater detail Color detector unit 500.

Color detector unit 500 includes two photoelectric detectors (photoelectric detector 505 and photoelectric detector 515), light Source 502, opaque barrier 510 and opaque barrier 511.Light source 502 can be any suitable light source according to the disclosure.Such as Shown in figure, light source 502 includes red, blue and green light source, but different number and class can be used according to different examples The light source 502 of type, this can permit light source and can be opened and closed, that is, independent chopping.Independent chopping is for emitting not Light source 502 with color allows color detector unit 500 that light output is customized to the light of specific wavelength.For example, particular color Change indicators can photoresponse to the specific wavelength of specific pH level it is quicker.

Photoelectric detector 505 and 515 can be any suitable photoelectric detector according to the disclosure.Photoelectric detector 505 and 515 are connected to processor 206.Separation distance 520 and light source 502 between light source 502 and photoelectric detector 505 with Separation distance 521 between photoelectric detector 515 can be adjusted based on application.Particularly, light source 502 and photoelectric detector 505 or 515 closer to together, and the part of photoelectric detector from the received light of light source 502 is more (and to be connect from environment light 250 The part of the light of receipts is fewer).Only as an example, separation distance 520 and separation distance 521 can be in the separations of 0.1mm to 2mm It adjusts.Other distances and configuration are also possible.As distance increases, in the identical situation of every other condition, photoelectricity inspection Device is surveyed to reduce from the intensity of the received light of light source.In addition, measured focus area increases as distance increases.With distance Reduction, sensor focuses more on the smaller area immediately below sensor.

As shown, using two photoelectric detectors 505 and 515.Photoelectric detector 505 and 515 may be positioned such that each other In parallel.In the configuration, the combination of photoelectric detector 505 and 515 provides output letter more stronger than other situations to processor 206 Number.Additionally provide an advantage using more than one photoelectric detector, i.e., if sensing system relative to object (for example, Color-bar 351) misalignment, then it can reduce error.

Color detector unit 500 may include one or more between light source 502 and photoelectric detector 505,515 A opaque barrier 510-511.Opaque barrier 510-511 reduction is propagate directly to photoelectric detector 505 from light source 502 and does not have There is the light quantity gone out from object reflection.Opaque barrier 510-511 can be rubber mat or similar material.On the one hand, Photoelectric Detection Device 505 or 515 may include the opaque outer of following such opaque barrier or photoelectric detector 505 or 515 can be with Following manner compacting: i.e. opaque outer is between LED and photodiode.On the one hand, opaque barrier is omitted 510-511 is to simplify design.

Fig. 6 is the flow chart for describing the method for the detection color according to some aspects of the disclosure.Fig. 1 or Fig. 2 will be combined The color detection exemplary methods that describe Fig. 6 using 111；However, according to different examples, it can be using appointing according to the disclosure What suitable Color Detecting System.In addition, can be executed by external equipment in conjunction with Fig. 6 operation described, such as via wirelessly connecting It is connected to the monitor equipment or external server of baby's sensing system 100.

At the box 601 of method 600, color detection obtains the of received environment light from photoelectric detector using 111 One measurement.Photoelectric detector 204 detects the existing environment light simultaneously intensity of the expression of the color of output light or existing broad-spectrum light Expression.For example, photoelectric detector 204 can produce the electricity output proportional to the wavelength or intensity of light is received.On the one hand, Photoelectric detector 204 can provide three outputs, and each output corresponds to red, green or blue: first with receive in light Red amplitude it is proportional, second is proportional with the amplitude of green received in light, the blue in third and reception light Amplitude it is proportional.

Photoelectric detector 204 provides the first measurement of light to processor 206.In this example, it is taken when light source 202 is closed The first light measurement is obtained, and the first light measurement indicates the environment light reflected from object 251.First light measurement can indicate broad-spectrum light Intensity.

The step of method 600, can be by the sensor encapsulation 400 of the Fig. 4 placed in the absorbent commodity of combination Fig. 3 description It executes.Because sensor encapsulation 400 may include one or more color detector unit 420a-n, exists be more than one wherein In the one side of a color detector unit 420a-n, sensor encapsulation 400 can measure environment light at multiple photoelectric detectors Level.Photoelectric detector in each color detector unit 420a-n can independently execute step 601-605.

At the box 602 of method 600, color detection makes light source emit light on object using 111.More specifically, place It manages 206 activating light source 202 of device and reaches scheduled pulse interval.In this example, baby's sensing system 100 only includes a light Source 202.But in some instances, multiple light sources can simultaneously or separately chopping.For example, using sensor encapsulation 400 Various aspects may include more than one color detector unit 420a-n.Light in each color detector unit 420a-n It source can individually chopping or the chopping together with other light sources.

At the box 603 of method 600, color detection obtains the second survey from photoelectric detector during transmission using 111 Amount, the second measurement include environment light and the transmitting light from object reflection.Processor 206 is in the pulse-on from light source 202 The second light measurement is obtained during time interval.Second measurement includes environment light and the light from light-pulse generator 202.Such as sensing In the aspect of device encapsulation 400, the photoelectric detector in each color detector unit 420a-n obtains the second light measurement.Color Sensing determines the color of object using 111 using the first measurement and the second measurement.

On the one hand, color detection can obtain more than one using 111 there are environment light and pulsed light Measurement.Multiple measurements can be averaged together to form the single measurement that may be used as the second measurement by processor 206.

At the box 604 of method 600, color detection is using 111 by being removed from the second measurement based on the first measurement Ambient light signal determines the AVHRR NDVI of reflected light.Removal can be executed in analog domain or numeric field.

For example, processor 206 can remove the first light measurement by filtering in analog domain from the second light measurement.Example Such as, processor 206 is subtracted from the second measurement that expression environment light is combined with the reflected light from light source 202 indicates environment light First measurement.Subtraction the result is that from the object 251 of such as color change indicator reflect light.

Processor 206 can operate in the digital domain.For example, processor 206 by the first measurement be converted to red, green and The number or numerical value of blue level indicate.Processor 206 by the second measurement be converted to red, green and blue level number or Numerical value indicates.Processor 206 calculates new red level by subtracting the first measurement from the red level of the second measurement, passes through The first measurement is subtracted from the green level of the second measurement to calculate new green level, and passes through the blue electricity from the second measurement It is flat to subtract the first measurement to calculate new blue level.New red, green and blue level are indicated from the light of object reflection Color.

At the box 605 of method 600, color detection determines (i) body exudates based on AVHRR NDVI using 111 In the presence of or (ii) existing for body exudates volume in one.Processor 206 exports the color of object and proposes the color Supply microcontroller 101.The color detection executed on microcontroller 101 receives color value from processor 206 using 111, and The presence of body exudates is determined using the data structure of such as table or indicates the value of body exudates volume.Microcontroller Device 101 can store several tables, for example, a table is convenient for the color on the color change indicator of such as bromocresol green It is mapped to pH level, and another table is convenient for the measurement for the volume that color change indicator is mapped to body exudates or is deposited In.

In addition, as discussed, color detection can execute color calibration using 111.Color detection can be incited somebody to action using 111 Red, green and blue level conversion is tone, saturation degree and brightness/value, and is executed to tone, saturation degree and brightness/value It calculates.Color calibration can be realized via table.For example, for red, green and blue given triple, according to a certain amount of Adjust these values.Color calibration can also execute in different domains, such as tone, saturation degree and brightness or tone, saturation Degree and value.

On the one hand, color detection can deposit the presence for determining body exudates in the case of movement using 111.Example Such as, while color detection is carrying out measurement using 111, baby makes sensor encapsulation 400 mobile.In such case Under, the known spirit of absorbent commodity or color-bar under the light of two or more different wave lengths can be used in color detection application Sensitivity determines the presence of effluent.In one example, even if in the presence of motion, color detection applies 111 It can detecte the response being greater than to the response of feux rouges to blue light.

On the other hand, color detection can detecte when absorbent commodity is not attached to baby using 111.In this feelings Under condition, transducer sensitivity changes to threshold value is lower than, this is by color detection using 111 detections.

Fig. 7 is the process for being used to determine movable illustrative methods from movable sensor according to some aspects of the disclosure Figure.Method 700 can be realized by activity classification using 116.In addition, can be executed by external equipment in conjunction with Fig. 7 operation described, Such as via the monitor equipment or external server for being wirelessly connected to baby's sensing system 100.

At the box 701 of method 700, it includes when being directed to one group that activity classification, which is received using 116 from movable sensor 130, Between each of section inertia measurement time series data.Inertia measurement may include acceleration or angular speed.For example, accelerating Degree meter can provide the triple numerical value corresponding to the direction x, y and z.Activity classification using 116 periodically to accelerometer into Row sampling is with generation time sequence data.Processor 206 marks each triple with timestamp, generate include sensor measurement and Pair of timestamp.Activity classification can also periodically sample gyroscope using 116.In conjunction with the survey from accelerometer Data are measured, activity classification can be determined using 116 including gyroscope measurement (such as angular speed), accelerometer measures (such as x- The triple of y-z value) and timestamp one group of data.

On the one hand, activity classification analyzes measurement data using 116 in real time, and can be with real-time update activity measurement functions Or prediction model.Alternatively, activity classification can once analyze sample block using 116.For example, activity classification application sequence 116 can be with It is described right to buffer, pair until number of thresholds has been received, the then movement on analysis time window.

At the box 702 of method 700, activity classification using 116 according to the subset of time series data, according to from inertia It measures derived statistical data and calculates activity function.Statistical data may include such as statistical variance of (i) inertia measurement or (ii) The data of the root mean square of inertia measurement.Activity classification using 116 using activity measurement functions to determine activity level.It can export Different activity measurements.For example, statistical variance, standard deviation or root mean square that activity classification can calculate signal using 116 (RMS).The measurement that activity classification can use another customization using 116 based on accelerometer or gyro data.For example, can make Calculate customization measurement to minor function, customization measurement level of quantization activity A for given n sample of quantity, wherein Sx, Sy and Sz is respectively the summation of the difference of two squares of each mean value in x, y and z dimension:

At the box 703 of method 700, activity classification is greater than the first threshold based on the measurement from activity function using 116 Value and it is less than second threshold and determines the activity indicated by time series data subset.Activity classification can be based on tool using 116 There is the activity level of certain value range to determine such as sleeping or awake etc activity.For example, if activity function measurement activity Level is lower than first threshold but is higher than zero, then activity classification determines that baby is in hypophypnosis using 116.If activity function is surveyed It measures mobile and horizontal and is lower than lower second threshold, then monitor application determines that baby is in deep sleep.Activity classification applies 116 State machine can be used to determine active state.

As discussed, on the one hand, in addition to or instead of algorithm or state machine, activity classification can be used pre- using 116 Model is surveyed to determine the activity of baby.Activity classification applies 116 by accelerometer measures, gyroscope measures or activity measurement functions Output be supplied to prediction model.Prediction model discussed here can be machine learning model, such as decision tree classifier or Regression model.Other models are also possible.Being trained to prediction model with the wearer for determining sensor is in left-hand side Feeding, in right-hand side feeding, sleep, awake and play, embraced or be seated in its back.Other detectable activities can be with Including seat, plays, creeps, walks.Activity classification can provide the number of one or more periods using 116 for prediction model According to.In this way, prediction model can determine activity based on the activity level of current or past.

Fig. 8 is according to some aspects of the disclosure for being lived by using prediction model to be determined according to movable sensor The flow chart of dynamic illustrative methods.In addition, the operation about Fig. 8 description can be executed by external equipment, such as via wireless It is connected to the monitor equipment or external server of baby's sensing system 100.

At the box 801 of method 800, it includes being directed to one group of time that activity classification, which is received using 116 from movable sensor, The time series data of the inertia measurement of each of section.At box 801, activity classification is received using 116 and is generally tied Close time series data described in box 701.

At the box 802 of method 800, activity classification calculates such as (i) inertia according to time series data using 116 and surveys The activity function of the root mean square of the statistical variance of amount or (ii) inertia measurement.At box 802, monitor is using generally The activity measurement functions in conjunction with described in box 702.

At the box 803 of method 800, activity classification using 116 by the statistical variance of activity function (i) inertia measurement or (ii) root mean square of inertia measurement is supplied to trained prediction model to identify the activity of effort scale.More specifically, activity point The output of sensor measurement or activity function is supplied to prediction model using 116 by class.

Prediction model is trained to determine activity according to the mobile measurement of instruction.It can be by providing instruction to prediction model Practice data and adjusts the inner parameter of model iteratively to train prediction model to reduce error.Training data may include movement Pair of data and brass tacks (identified activity).For example, training data may include one group of movement and training data label " sleep " or " feeding ".Then, prediction model study is associated with activity by specific Move Mode.

Trained prediction model determines which classification activity belongs to from predefined one group of classification based on its training.Activity class Other exemplary lists be included in left side feeding, right side feeding, sleep, regain consciousness but play, embraced and be seated in back.Its He trains classification also possible.For example, prediction model can be trained to distinguish deep sleep and hypophypnosis, and area each other It point such as creeps, roll, evening sleeps, the activity of feeding or nursing etc.For example, activity classification may include being trained to using 116 To distinguish the prediction model of sleep, awake, excited (stir) or stable state, and it is trained to distinguish hypophypnosis and depth Another prediction model of sleep.It is excited to indicate the mobile state more than first threshold amount of baby, and stably expressed baby is The tranquil and mobile state for being less than second threshold amount.

At the box 804 of method 800, activity classification applies 116 sons from prediction model reception and time series data Collect corresponding movable determination.For example, prediction model from such as left-hand side feeding, right-hand side feeding, sleep, it is awake and 116 prediction is applied in an offer in the training classification that its back plays, embraced or is seated to activity classification.

Fig. 9 is the side of the volume of the body exudates in the detection absorbent commodity described according to some aspects of the disclosure The flow chart of method 900.From diaper load using 102, activity classification applied using 116 and color detection 111 interpretation side Method 900, it will be appreciated that the different step of method 900 can be executed by these or other application.In addition, an application All steps can be executed.In addition, can be executed by external equipment in conjunction with Fig. 9 operation described, such as via being wirelessly connected to The monitor equipment or external server of baby's sensing system 100.

At the box 901 of method 900, color detection is received when light source is closed from photoelectric detector using 111 Environment light first measurement.At box 901, color detection is executed using 111 and is described with the box 601 about method 600 Similar functions.

At the box 902 of method 900, color detection obtains while source emissioning light from photoelectric detector using 111 Obtain the second measurement.Second measurement includes the measurement of environment light and the transmitting light reflected from absorbent commodity.At box 902, face Color sensation, which is surveyed, executes the similar functions described with the box 601-602 of combination method 600 using 111.

At the box 903 of method 900, diaper load is using 115 by being removed from the second measurement based on the first measurement Ambient light signal come determine from absorbent commodity reflect light AVHRR NDVI.At box 903, diaper load applies 115 Execute color detection and similar functions described in the box 604 of combination method 600 using 111.

At the box 904 of method 900, diaper load, which determines in absorbent commodity using 115 according to AVHRR NDVI, to be urinated The presence of liquid.At box 904, diaper load makes color detection using the side of 111 execution and combination method 600 using 115 Similar functions described in frame 605.

At the box 905 of method 900, diaper load applies 115 full levels for determining absorbent commodities.Full level It reflects relative to can be filled by the absorbability article for the memory space total amount that body exudates are filled or absorb body Amount of memory in the absorbent commodity of effluent.In some cases, the full level of absorbent commodity is from diaper Derived from existing volume of urine.Different inputs can be used to determine the volume, the input all (i) in this way is deposited from urine The state (for example, awake or sleeping) of elapsed time, the time of (ii) more diaper-exchanging and (iii) baby since.

Diaper replacement can indicate by nursing staff, for example, via user interface or to baby's sensing system other are defeated Enter.Alternatively, diaper load using 115 can by detection sensor from baby diapers removal or measurement it is wet The reduction of degree is not present to detect the presence of new diaper.

Understand that the time quantum under sleep and/or waking state helps to improve prediction.For example, relative to daytime, Ying Erke It can be urinated during night with slower frequency and quantity.In addition, having more accurately prediction, baby's sensing system in this respect With additional benefit, i.e., if there is an urgent need to more diaper-exchangings, nursing staff is allowed to sleep more long.As about Fig. 7-8 institute Explain, activity classification using 116 movable sensor can be used in conjunction with prediction or state model determine whether baby sleeping, Awake, rest etc..Therefore, at box 905, exemplary operations include the operation executed in method 700 and/or 800.

In addition, in some cases, diaper load receives specific type, the brand about used diaper using 115 Or size (for example, standard size, such as 1,2,3 etc.) or diaper are the inputs of conventional (daytime) diaper or diaper overnight.Every Enuresis nocturna piece can have bigger absorbability.In addition, diaper load can receive the demographics letter about baby using 115 Breath, such as age, gender, weight etc. may be used as the basis of prediction.For example, biggish baby may urinate more, cause Need smaller baby more diaper-exchanging earlier.

Diaper load can also determine the time until absorbent commodity is full of using 115.For example, diaper load application The capacity (for example, the volume that theoretically can store the liquid in diaper) of 115 access diapers, calculates the increased rate of volume (for example, frequency and amount of the urine event since based on more diaper-exchanging) simultaneously calculates the time that diaper will be filled with.Diaper load is answered Baby's sensing system 100 can be made to send alarm on the time for reminding nursing staff to look after baby or before the time with 115.

Diaper load can be pre- to establish diaper based on the specific dimensions of diaper, diaper type or other parameters using 115 Phase keeps the model of how much liquid.Then, based on this model, the application of diaper load can based on one or more event evaluation it is special Determine the load of diaper at any time.For example, application it is contemplated that (1) when with new diaper replacement diaper (for example, currently diaper baby Dressed how long), (2) diaper whether moist (or when diaper moisturizes for the first time), (3) since diaper is got wet Other data of the time it takes or (4) such as diaper type under sleep or waking state.

Statistical method can be used.Diaper load using 115 can use one or more regression models (for example, it is linear, Two is inferior) or machine learning model (for example, decision tree classifier or other disaggregated models) solve function.By solving mould Type, the volume that diaper load determines urine present in diaper using 115.

Figure 10 is described according to some aspects of the disclosure for executing and the detection of color detection and body exudates The figure of the exemplary computing system of relevant function.The some or all of components of computing system 1000 may belong to the microcontroller of Fig. 1 Device 101 or processor 206.For example, color detection can be operated using 111 in computing system 1000.Computing system 1000 is wrapped Include the one or more processors 1002 for being communicatively coupled to one or more memory devices 1014.Processor 1002 executes meter (it can be stored in the non-transitory computer executable instructions in memory devices 1014 to calculation machine executable program code Form), access the information or both of the above being stored in memory devices 1014.The example of processor 1002 includes micro process Device, specific integrated circuit (" ASIC "), field programmable gate array (" FPGA ") or any other suitable processing equipment.Processing Device 1002 may include the processing equipment of any quantity (including one).

Memory devices 1014 include any suitable computer-readable medium, such as electronics, light, magnetism or other can Computer-readable instruction or the storage equipment of other program codes are provided to processor.The non-limiting of computer-readable medium is shown Example includes disk, memory chip, ROM, RAM, ASIC, optical memory, tape or other magnetic memories or processing equipment can It is read from any other medium of instruction.Instruction may include by compiler or interpreter from any suitable computer volume The processor specific instruction for the code building that Cheng Yuyan writes, suitable computer programming language include such as C, C++, C#, Visual Basic, Java, Python, Perl, JavaScript and ActionScript.

Computing system 1000 can also include multiple outside or inside equipment, such as input or output equipment.For example, calculating System 1000 is shown with input/output (" I/O ") interface 1010, can receive input from input equipment or set to output It is standby that output is provided.Bus 1006 also may include in computing system 1000.Bus 1006 can be with communicatively coupled computing system 1000 one or more components, and allow the communication between these components.

Computing system 1000 executes program code, and the program code configuration processor 1002 is above in conjunction with Fig. 1-5 to execute One or more operations of description.Color detection applies for 116 program generation using 111, diaper load using 115 or activity classification Code can be the form of non-transitory computer executable instructions, and it is in memory devices 1014 or any suitable to may reside within Computer-readable medium in, and can be executed by processor 1002 or any other one or more suitable processor. It executes such program code configuration or (multiple) processor is made to execute the operation described herein in regard to microcontroller 101.Attached Add or alternative aspect, above procedure code can store can from remote storage device via data network in computer system 1000 In one or more memory devices of access.Memory devices 1014 can be used in microcontroller 101 and any process.Storage Device equipment 1014 can store such as appendage or application (such as the color detection application by executing on processor 1002 111) data used.

Computing system 1000 can also include at least one network interface 1004.Network interface 1004 includes being suitable for establishing To any equipment or equipment group of the wired or wireless data connection of one or more data networks.The non-limit of network interface 1004 Property example processed includes ethernet network adapter, WiFi network, bluetooth or low-power consumption bluetooth (BLE), modem etc..It calculates System 1000 is able to use network interface 1004, and via data network and one or more, other calculate equipment or computer-readable Data Source communication.

There is set forth herein many details to provide the thorough understanding to theme claimed.However, this field The skilled person will understand that theme claimed can be practiced without these specific details.In other examples Under, it is not described in method, apparatus or system known to persons of ordinary skill in the art, in order to avoid obscure master claimed Topic.

Unless expressly stated otherwise, otherwise it should be understood that throughout the specification, utilizing such as " processing ", " calculating ", " fortune Calculate ", the discussion of the terms such as " determinations " and " identification " refer to the movement or process of calculating equipment, calculating equipment be, for example, one or Other letters of multiple computers or similar electronic computing device, manipulation or conversion in memory, register or computing platform The data of physical electronic or the quantity of magnetism are expressed as in breath storage equipment, transmission equipment or display equipment.

One or more system discussed here is not limited to any specific hardware structure or configuration.Calculating equipment can wrap Any suitable component arrangement is included, the result with one or more inputs for condition is provided.The suitable equipment that calculates includes visiting Ask computer system of the multipurpose based on microprocessor of the software of storage, the software of the storage is by computing system from general-purpose computations Device programs or is configured to the dedicated computing device for realizing the one or more aspects of this theme.Any suitable volume can be used The combination of journey, script or other kinds of language or language come for program or the software of configuring computing devices in realize this paper The introduction for including.

Some exemplary foregoing descriptions are presented merely for the purpose of illustration and description, and is not intended to exhaustion or incites somebody to action The disclosure is limited to disclosed precise forms.In the case where not departing from the scope of the disclosure and the accompanying claims, ability Field technique personnel will be clear that its many modification and reorganization.

Claims

1. a kind of method for the load for predicting wearable absorbent commodity, which comprises

The first measurement of received environment light is obtained from photoelectric detector when light source is closed；

Obtain the second measurement from the photoelectric detector when source emissioning light, second measurement include the environment light and from The measurement of the transmitting light of absorbent commodity reflection；

Ambient light signal is removed from second measurement by the first measurement based on the environment light, is determined from absorbability object The AVHRR NDVI of the light of product reflection；

The presence of body exudates in the absorbent commodity is determined according to the AVHRR NDVI of the light；

According to including time series data to the inertia measurement of each of one group of period, active state is determined；

The full level of the absorbent commodity is determined, wherein the determination determines depositing for the body exudates based on (i) certainly Elapsed time and (ii) described active state since.

2. according to the method described in claim 1, further include:

Being solved by using regression model, there is the function for the input for indicating the elapsed time and the active state to count Calculate the time until the expection absorbent commodity is full of；With

On the time of calculating or before export recommended replacement described in absorbent commodity notice.

3. method according to claim 1 or 2, wherein determine ruler of the full level based on the absorbent commodity It is very little.

4. according to the method in any one of claims 1 to 3, wherein determine that the full level includes that (i) is used to return One or more of model or (ii) machine learning model indicate the elapsed time and the moving type to solve to have The function of the input of state.

5. method according to claim 1 to 4, wherein determine the full level be also based on (iii) since The absorbent commodity is placed on the time additionally gone through risen with baby.

6. the method according to any one of claims 1 to 5, wherein the active state includes that (i) is sleeping or (ii) is clear One in waking up.

7. method according to any one of claim 1 to 6, wherein the absorbent commodity is the diaper of baby's wearing, And wherein the method also includes accessing the lineup mouthful statistics about baby, and wherein determine the full level part Ground is based on the demographics.

8. method according to any one of claim 1 to 7 further includes the type for accessing the absorbent commodity, wherein The type is (i) night diaper or (ii) one of diaper in the daytime, and wherein determines the full level part of the foundation In the type.

9. method according to any one of claim 1 to 8, wherein determine that the active state includes:

Multiple three-dimensional inertia measurements are received for a period of time from movable sensor；

Calculate the statistical data derived from the inertia measurement；

The multiple inertia measurement and the statistical data are supplied to prediction model；And

Determining activity is received from the prediction model and based on the inertia measurement.

10. method according to any one of claim 1 to 9, wherein second measurement includes feux rouges, green light and indigo plant The independent measurement of light, wherein determine that the AVHRR NDVI of the light includes the independent intensity of determining feux rouges, green light and blue light, and Wherein determine intensity of the full level based on feux rouges, green light and blue light.

11. method according to any one of claim 1 to 10, including based on the absorbent commodity to different wave length The wavelength for the light that the sensitivity selection of light is emitted by the light source.

12. method according to any one of claim 1 to 11, wherein the absorbent commodity includes becoming containing color Change printing or the coating zone of indicator.

13. method according to any one of claim 1 to 12, further includes:

The color calibration value of storage is obtained from memory；And

The color calibration value of AVHRR NDVI and storage based on the light determines the color of absorbent commodity, wherein using white Object determines stored color calibration value, and the determining color includes white level correction.

14. a kind of system for determining the volume of the body exudates in absorbent commodity, the system comprises:

Light source；

Photoelectric detector；

Movable sensor；With

Processor is configured as:

The first measurement of received environment light is obtained from the photoelectric detector when light source is closed；

It is determined by removing ambient light signal from second measurement based on the first measurement of the environment light from absorbability The AVHRR NDVI of the light of article reflection；

The time series including the inertia measurement for each period in one group of period is obtained from the movable sensor Data；

To at least one of the AVHRR NDVI of external equipment transmission (i) described light or (ii) described time series data；With And

The full level of the absorbent commodity is received from the external equipment.

15. system according to claim 14 further includes the external equipment, wherein the external equipment is configured as When receiving the AVHRR NDVI of (i) described light and at least one in (ii) described time series data, based on from detecting According to the determining body exudates of the AVHRR NDVI of the light there are elapsed time and according to the time series number The full level is determined according to determining active state.

16. system according to claim 15, wherein the external equipment is additionally configured to ask by using regression model Solving, there is the function for the input for indicating the elapsed time and the active state to calculate until the expection absorbability object The time that product are full of.

17. system according to claim 15 or 16, wherein the external equipment be additionally configured to based on (iii) since The time additionally gone through that the absorbent commodity is placed on baby determines the full level.

18. a kind of system, comprising:

Computer-readable medium stores non-transitory computer executable instructions；And

Processing equipment is communicatively coupled to the computer-readable medium, executable for executing the non-transitory computer Instruction, wherein executing the non-transitory computer executable instructions configures the processing equipment to execute according to claim 1 To method described in any one of 13.